D, N, RD and RN differed significantly in their sleep lengths (F3, 33=31.63; P < 0.0001, one-way ANOVA followed by post hoc Tukey's multiple comparisons test), but D and RD (P = 0.925) or N and RN (P = 0.434) did not. Significant differences marked sleep lengths of RD from RN (t6 = 7.224, P < 0.0005). One-way ANOVA did not reveal significant differences between age (F3, 30=0.475, P = 0.702), body mass index (F3, 30=1.81; P = 0.333), work experience (F3, 30=0.41; P = 0.746) and pulse rate just after waking up (F3, 30=1.16; P = 0.341) volunteers working in different shifts. Volunteers had either morning or intermittent chronotype.
Figure 1A, C shows CV rhythm in D. Higher BP at CT12–13 (SBP-141.5±6.9 mmHg; DBP-99.08±8.4 mmHg) was followed by a night dip at CT18 (SBP-98.2±3.2 mmHg; DBP-58.3±2.3 mmHg), i.e., within 3 hours of sleep onset in day workers. A similar bimodal pattern was missing in N, who exhibited acrophase in SBP (CT4; 152.5±6.3 mmHg) and DBP (CT17; 95.3±8.7 mmHg) at different times, but the dip in BP occurred during nap (Fig. 2A and B) at CT11–12 (SBP-111.5±6.4 mmHg; DBP-66.8±5.2 mmHg). Student’s t-test revealed a significant difference between SBP (t46 = 1.782, P < 0.05) of D and N but not in DBP (t46 = 1.06, P = 0.15). Figure 1B, D shows CV rhythm in RS. No conspicuous peak could be observed in SBP (Fig. 2B) and DBP (Fig. 2D) of RS during RD and RN. A night dip at CT18 in SBP among RD (97.3±7.9 mmHg) and at CT23 in RN (113.1±6.2 mmHg) was observed, while there was a night dip at CT19 in DBP among RD (59.5±6.5 mmHg) and at CT23 in RN (67.7±6.1 mmHg). Student’s t-test revealed a significant difference between DBP (t46 = 1.78, P < 0.05) of RD and RN but not in SBP (t46 = 1.53, P = 0.06). However, hours from wake up, i.e., time of physiological day, significantly contributed to BP variations among both D vs. N (SBP- F23, 414= 4.117, P < 0.0001; DBP- F23, 414=2.516, P < 0.001) and RD vs. RN (SBP only F23, 345= 2.56, P < 0.001), as revealed by two-way ANOVA. Body temperature dipping of 1.5°C occurred in D prior to night sleep, whereas N exhibited greater dipping, i.e., up to 2.4°C, that conformed to naps during work hours. Temperatures dipping in RS did not conform to sleep, wake, or nap times; there was an interchange between nap-dipper and night-dipper status during rotating shifts in some volunteers. There were significant differences (F3, 92=49.43; P < 0.0001, one-way ANOVA followed by Tukey’s multiple comparison test) among the four groups. RS (RD and RN) experienced a significant (t23 = 8.698, P < 0.0001) change in body temperature during the shift transition. The mean temperature of RD and RN was higher than that of the corresponding control groups, i.e., D and N (RD-D = 0.89±0.04; RN-N = 1.2±0.23).
Figure 2 shows that acrophase of systolic and diastolic BP (geometric mean ±SE) occurred in D controls on CT8 and CT7 and in N on CT and CT10.25, respectively. RS exhibited systolic and diastolic BP acrophase at CT7 and CT7.25 hrs during the dayshift that changed to CT9.75 and CT8.6, respectively, during the night shift.
Night and rotational shift work dampened the amplitude of daily rhythm of blood pressure compared to controls. Temperature acrophase (Fig. 2C) occurred at CT-7.48 in D but much later in N (CT-13.45). RS had a higher body temperature that peaked at CT10.4 in RD and CT9 in RN. While a temperature dip in D existed in the initial hour of sleep onset (CT17.2), it was preponed in shift workers (N- CT10.75, RD- CT9, RN- CT14.2). The dynamic changes in additional CV parameters measured by ABPM, i.e., pulse, MAP and double product of BP, exhibited similar daily trends as SBP and DBP in the D (F96, 1249=1.203; P = 0.09), N (F96, 1000=0.7044; P = 0.98), RD (F96, 750=0.81; P = 0.89) and RN (F96, 1125=0.78; P = 0.94) groups.
All volunteers exhibited at least a 10% dip in temperature during the day, although few of them exhibited a 20% dip. All D exhibited a night dip (Nitdip) before sleep, which was used as a reference state for comparison in shift workers. Some N and RS also exhibited dipping during naps in work hours (Napdip). A total of 62.5% of N and 44.4% of RN exhibited Napdip, and 50% of RS changed between the Nitdip/Napdip statuses between the day and night shifts (Fig. 3). Daily dip in temperature positively correlated with self-reported quality of night sleep in RS but not in D and N.
A one-way MANOVA revealed statistically significant differences among the changes in systolic, diastolic and temperature rhythms (all shift-work: Wilks' Lambda = 0.094, F (5, 50) = 15.48, p ≤ 0.0001; N: Wilks' Lambda = 0.055, F (5, 40) = 3.26. p ≤ 0.0001; RD: Wilks' Lambda = 0.5, F (5, 30) = 3.9. p ≤ 0.0001 and RN: Wilk's Lambda = 0.30, F (5, 45) = 6.9, p ≤ 0.0001) with day workers.